The Five Engines Of eCRM.

Understanding Electronic Customer Relationship Marketing

Companies understand that Electronic Customer Relationship Marketing (eCRM) has significant potential, but they face the challenge of building the required technology infrastructure quickly and cost effectively. A knee-jerk reaction is to buy off-the-shelf applications, cobble together a database of web traffic and online purchase information, and launch an eCRM initiative. Unfortunately, many such efforts have met with poor results. Recent research indicates that 39% of online shoppers failed in shopping attempts and a staggering 66% of loaded online shopping carts were abandoned before the checkout process. Less than 5% of unique visitors become customers.

A more sound approach is to install a comprehensive software platform of five engines that together enable the eCRM business process. These five engines are:

1. The Customer-centric Information Store: To consolidate information about millions of customers together with preferences, permissions, and information that may be useful to them.

2. The Analysis and Segmentation Engine: To leverage this customer information to build a business campaign strategy and evaluate its success.

3. The Personalization Engine: To personalize the entire customer experience, configuring unique sets of messages and offers to each customer.

4. The Broadcast Engine: To proactively deliver information and offers to every customer via the media of his or her choice.

5. The Transaction Engine: To facilitate the interactions between customer and the company, either exchanging information or driving transactions.

Properly configured, these five engines collectively form a robust, scalable, and flexible platform for eCRM. Prefabricated and custom-made software can be seamlessly integrated into the platform to provide a virtual shopkeeper to millions of customers. Equipped with such infrastructure, companies can continually create significant customer value at Internet speed, automating the who, what, when, where, and how of sales and marketing.

Engine 1: The Customer-centric Information Store

eCRM initiatives depend on a 360-degree customer view. A Customer-centric Information Store integrates data from disparate information sources such as web sites, transactional systems, operational databases, call centers, enterprise resource planning systems, and third party data. This engine enables companies to recognize and respond accurately to customers, whether they purchase products through a physical store; telephone a call center; or browse a web site. Consequences of an incomplete customer view are often harsh. Presenting customers with inappropriate offers that dilute loyalty and trust stall eCRM efforts.

Developing a Customer-centric Information Store requires large-scale integration of disparate data from multiple sources. Three factors are critical to its success:

* Scalability: Companies tend to underestimate the volume of data that is required for developing a comprehensive 360degree view of customers and visitors. Clickstream data alone can consume several gigabytes every day. Transactional data, third party, and ERP data add exponentially to the volume. Detailed, atomic level capture and retrieval of every customer's profile, preferences, and transaction history places additional capacity demands on the technology architecture. Companies must plan for terabyte-sized Customer-centric Information Stores.

Flexibility: The Customercentric Information Store must accommodate multiple data models and database architectures and allow for integration with other back-end information systems. Without this flexibility, the usefulness of this engine will diminish over time. Customer information stores are dynamic, growing entities that have to keep up with every customer's interaction with the company.

* High Performance: Speed and accuracy of access to customer-centric information is essential for enabling a true value exchange with customers. Likewise, the ability to aggregate information at differing levels of abstraction (e.g., transaction, customer, and zip code) makes it possible to discern patterns of customer behavior (e.g., sizes of clothing sold by zip code). In automotive engineering, sports car engines provide the ability to drive at high speeds, turn around tight corners precisely, and deliver flawless performance on bylanes, country roads, or highways. Similarly, a high performance Customer-centric Information Store provides accurate information that can be captured in any part of the information store at any level of granularity with maximum flexibility and responsiveness to demand.

Effectively applied, the Customer-centric Information store decreases customer attrition rates and improves retention. Over time, the Customer-centric Information Store forms an inventory of customer trust and loyalty, becoming a valuable corporate asset and giving an enterprise a source of competitive advantage.

Engine 2: The Analysis And Segmentation Engine

Building trusted customer relationships depends on accurate customer segmentation. The Analysis and Segmentation Engine performs business analysis, segmentation, and prediction so that customer interactions take place in an appropriate and personalized manner. Without this engine, eCRM lacks the intelligence to be effective--even if it has massive volumes of customer-centric information.

There are three major categories of analysis and segmentation techniques: Online Analytical Processing (OLAP), Data Mining, and Statistics. Briefly, OLAP tools perform complex queries on a database, Data Mining tools discover unforeseen associations using pattern-matching algorithms, and Statistical tools perform complex mathematical operations on sets of data. Each technique has its strengths.

No one technique covers the gamut of eCRM applications. eCRM requires a combination of Statistics, Data Mining, and OLAP solutions to perform rapid, accurate, and consistent customer segmentation. The new breed of analysis and segmentation is built on iterative, "Collaborative" ROLAP (Relational OLAP) tools and adds the best features of Statistical and Data Mining analysis techniques. Together, they iterate to a final answer to analytical questions. eCRM marketing campaigns are inherently iterative since they require ad-hoc questions about customer preferences, demographics, and transaction patterns. With Collaborative ROLAP, campaign formulation evolves through experimentation and discovery. Iteration takes place between a server, where data is stored, and the "middle tier" applications that are used for complex analytical operations. This approach is essential for extracting maximum benefit from an integrated Customer-centric Information Store. The hybrid will successfully hypothesize, target a sub-segment of customers, and verify results.

Engine 3: The Personalization Engine

New technology makes it possible to personalize products and services for large numbers of customers in a cost-effective manner by lowering the marginal cost of personalization. Until now, personalized attention and service were labor intensive and not scalable to serve a large customer base without high costs. Naturally, most companies provide personalized attention to a small group of selected clients who are "worth it." Today, every Amazon.com customer gets personal recommendations for books--in large part due to Amazon's personalization technology. The addition of thousands of customers to Amazon's base has a minimal marginal effect on their cost to sustain this high level of service. This "scalable" personalization is unprecedented in business history.

Achieving this level of one-to-one service requires a robust Personalization Engine. Engine 3 identifies appropriate messages and offers necessary tools to create valued one-to-one customer relationships with millions of customers. It builds customer profiles, enables customized products and services offerings, and fosters trusted relationships. The Personalization Engine leverages the Customer-centric Information Store to provide a "virtual storekeeper" who knows each customer, looks out for their needs, and knows when and how to reach them.

Classic approaches to personalization fall into three broad categories: Rules-based, Collaborative Filtering, and Inference Models. Customer Experience Personalization (CEP) utilizes all three approaches in conjunction with large volumes of information to create a uniquely holistic experience for each individual customer.

CEP is proactive and data driven. For example, regular National Weather Service reports on pollen count could be used in conjunction with patient profiles to alert specific asthma patients to renew their allergy medication. This is superior to personalization that solely relies on static information about customer transaction patterns. Customers and visitors dislike information clutter. Given a chance, they will clearly indicate their preferences for the type of information they like to receive, the periodicity of subsequent communication, and the device for reaching them. This information is a powerful tool to enhance the level of personalization of customer experience. For example, a customer's self-stated interest in receiving news on Federal interest rate changes can be leveraged by a bank to package news of an impending rate hike with an offer for a low-interest credit card or home equity loan. A good Personalization Engine should include a set of tools to create subscription services and interfaces tha t provide "in-a-box" capability.

Web-content personalization is necessary, but not sufficient, for CEP. A customer who responds positively to a cleverly worded e-mail may prefer brevity and clarity in voice mail messages. True personalization demands that each customer experience is rich and appropriately tailored by the type of device. This provides greater degrees of freedom to an eCRM initiative. Armed with a powerful Personalization Engine, a company can provide unique experiences to the same customer on different devices. An added benefit of this is formulation of an increasingly richerpicture of customer preferences for presentation and promotion.

Engine 4: The Broadcast Engine

Continuous customer interaction and value exchange require 24/7 access to customers. While the web is becoming increasingly pervasive, few customers are on-line all day. However, they are increasingly reachable through other communication devices. A recent META Group study projects that the average number of communication devices per consumer will quadruple in the next three years.

Growing appetite for non web communication is not just a consumer phenomenon, but a business shift, as well. Corporate use of wireless technology is increasing dramatically. Successful eCRM requires an engine that reaches millions of customers wherever they are: at home, via phone, or TV set-top box; at work; via e-mail; or on the road, via WAP phone or pager. A scalable Broadcast Engine that is built on an open architecture and supports all communication devices enables this level of customer interaction.

Characteristics Of A Broadcast Engine

* Multi-Media/Multi-Channel (MMMC) Capability: Static text e-mail has limited appeal and increasingly resembles "junk mail." Multi-media communication that combines graphics, video, and audio stimulates increased viewership. Reaching customers through the channels of their choice is even more important. An e-mail from an online brokerage informing a client of a 20% fall in a stock is useless if the customer is not online. Such information is best delivered to the customer's cell phone or pager. Multi-channel broadcasting of personalized information enables recipients to rapidly transform information into action. An intelligent Broadcast Engine identifies and delivers information through web, wireless, and by voice, using text-to-speech synthesis, and capitalizes on the media possibilities offered by each device.

* Open Architecture: Rapid technological change has brought new information sources, as well as communication and information transfer services. An open architecture allows for the integration of new data sources and transmission methods as they emerge (e.g., broadband media). In order to be open and flexible, the architecture should adhere to Internet protocols, have broad database support, and open, documented APIs. The open nature of the Broadcast Engine will allow it to keep pace with e-business as it continues to evolve.

* Scalability: The Broadcast Engine must be capable of reaching millions of users. Without large-scale reach, eCRM initiatives will stall as a business grows.

Broadcast Engine Benefits

* Integrated Channel Strategy: With multi-channel capability, companies can leverage each channel's unique benefits. Few customers would like constant interruptions on their cell phones giving them promotional material, but a crisp voice message relating a significant shift in stock prices would be appreciated. A tempting product offer in a well crafted, graphically rich format is best sent to a desktop PC e-mail box connected through a high bandwidth link. A Broadcast Engine enables an integrated and flexible communications strategy for each individual customer.

* Increased Customer Switching Costs: Competitors are a mouse click away on the web, so customers' costs of switching to other sites are low. Customers' switching costs increase as the number of channels of interaction with the company increases. Fewer customers are likely to shift to a new service provider if they have to renew e-mail addresses, cell phone numbers, and WAP phone alert service numbers. A Broadcast Engine with multi-channel capability increases the likelihood of customer retention.

* Speed to Transaction: The Broadcast Engine delivers anytime and anywhere. The likelihood of brokering or consummating a transaction increases dramatically when a company can reach the customer first. A video retailer can alert film buffs about new video releases through a voice mail or pager and secure transactions faster than stores that merely have web sites or storefronts. The Broadcast Engine accelerates the speed of a transaction cycle and allows a company to be "first to transaction."

Engine 5: The Transaction Engine

An effective Transaction Engine promotes information exchange between every customer and the enterprise. Like the small town shopkeeper who often conversed with his customers and remembered significant details, the Transaction Engine maintains customer contact and transmits information to the Customer-centric Information Store for later use. Leveraging the other four eCRM Engines, the Transaction Engine develops informed lifelong customer relationships.

The Transaction Engine facilitates the value exchange and provides a single interface to any set of information sources. The Transaction Engine also acts as a third-party purchase facilitator for the consumer. The consumer provides relevant preferences via any device and the Transaction Engine vends the relevant information. Smart consumers want trustworthy product information before every potential purchase. To gain confidence in their product purchases, customers will interact with many vendors to gather information, conduct comparative analysis, and then decide which products to buy. The Transaction Engine promotes and ultimately brokers customer transactions.

The Transaction Engine manages the flow of information and services through each customer device and provides appropriate value-added features and functionality. This is achieved by integrating closely with the Broadcasting and Personalization Engines.

Companies will have billions of interactive customer contacts a year. They have to avoid inundating customers with unwanted information and services. Through the utilization of subscription methods (via web, wireless, or voice) customers can sign up for unique interactive information services based on their own preferences. The Transaction Engine performs this function, keeps track of the customer subscriptions, and uses this information to initiate interactions with individual customers.

Without the customer's permission, a company is locked out of a potentially lucrative relationship. As part of the value exchange, customers grant companies the permission to interrupt the flow of their daily life with interactions. This is of significant value to the company as it represents the customer's trust and loyalty. These translate directly to superior economics for campaigns in terms of higher response rates and lower conversion costs. Permissioning capabilities are a core characteristic of The Interaction Engine.

Permission-based interaction allows enterprises to engage in learning relationships with customers. With every interaction with the customer, corporations learn more about particular customers' product, service, and contact preferences. In order for customers to receive comparable service from competitors, they would have to re-teach the competitor their preferences. This represents a switching cost for customers, which serves to enhance customer retention. The eCRM tsunami has begun and executives in every major company will have to formulate and execute an eCRM strategy rapidly in order to keep up with competition.

Mark LaRow is the vice president of the applications group at MicroStrategy, Inc.